JP2 vs JPG File Size/Quality

I recently discovered the JP2 format. I downloaded an extension that lets me import it into Adobe Photoshop CS3, and I also installed the software program Pixelmator. Now I’m just trying to understand how it compares to JPG in terms of quality and file size.

I thought JP2 was supposed to offer higher quality in a smaller file size, but when I export the same image @ 65% in both formats, the JPG is actually a smaller file size than JP2.

Actually, I should get straight to the point…I’m going to create an enormous library of images in the public domain, which I want to store in an image management system, Aperture for the time being. When I want to use an image on my website, I’ll retrieve it from Aperture, process it in Photoshop and save it as a JPG, GIF or PNG. If you were going to store 10,000 images in Aperture, what format (JPG and JP2) and quality would you choose?

Photoshop apparently saves JPG images at 60% quality by default, and they look pretty nice to me.

Thanks.

It’s not as simple as a pure quality percentage to size comparison - have a look at this analysis that shows at the same percentage (and a similar size) the quality at lower settings is superior in jp2. jp2 tends not to get ‘block’ artefacts as it works in a different way. The implication is that for equivalent visual quality you would be able to apply a higher compression ratio with jp2.

To throw you another curveball, another option is single frame h.264 which is supposedly even better, though both of these more modern formats suffer from lack of widespread adoption.

If I understand correctly, you want to create a repository of images? Well you should be using a file type that uses a lossless compression, JPEG is not one of them. JPEG2000, does provide lossless compression. As does, PNG, TIFF and others. You should worry less about file size and more with quality. Lossy compression of JPEG will be trouble some.

Thanks for the tips. Sounds like I definitely need to get away from JPG’s. I’ve never even heard of single frame h.264 before; I’ll have to check it out.